7 Reasons Science Says Political Polls Are Less Accurate Than A Broken Clock

04/05/2017 07:02 pm ET Updated Apr 06, 2017

This week, Donald Trump’s approval rating dropped to 34 percent... the lowest of any US president. Ever. If you don’t count the day the Russians released Polaroids of Gerald Ford getting peed on by a prostitute at a Motel 6 in Moscow.

HOWEVER: there’s a secret the media doesn’t want you to know: Polls are to science what Big Foot is to animals. Sure, everyone loves to hear about the latest big foot sighting. but, once you look closely, examine, and ask a few questions, you discover that Polls… like Bigfoot, are 100 percent total bullshit.

Polls offer a snapshot of public opinion. A group of people are asked a bunch of specifically designed questions to determine how they feel about one issue or another and then THOSE answers are meant to be illustrative of the opinion of the entire population.

But, unlike laws, polls have no tangible consequence. Plus, tomorrow, the numbers could be totally different. If Trump has a low approval rating he isn’t fined. Or flogged. Or fired. In fact, as Trump demonstrates, you get to keep your job, AND strip the arts budget to the point where if Elmo wants health insurance, he’s gonna have to double the number of hand jobs he’s giving behind the 7-11.

Even worse, and more importantly: polls are routinely wrong!

The night before the British went to vote on staying in the European Union or Exiting, the polls had “Stay” winning by 10 percent. When the votes were tallied, “Exit” won by 4 percent.

In the polls conducted during the run up to the 2016 Presidential Election, the results were so out outrageously out of line… Gretchen Carlson sued… and was paid 40 million in hush money. The last time this many polls were wrong… Hitler invaded Poland.

FIRST PROBLEM: The first problem with political polls is: The people giving their opinion are probably NOT a representative sample of voters. 20 years ago, polling companies used computers to randomly call land line phones… in people’s homes.

Back then, more than 98 percent of homes had phones. So, if you randomly dial 20-thousand homes… and just 7 or 8 percent of people picked up – with about 1600 people surveyed, you were likely to have reached a statistically significant audience.

Today, the number of homes that don’t have a land line phone is: 53 percent. And a preponderance of those that DO have a land line phone? They’re over the age of 60.

Poll companies also randomly call cell phones - a recent Gallup poll said 70 percent of people polled were on their cell phone… which means they’re likely younger… So, you’ve got a preponderance of opinions… from a group that traditionally doesn’t vote.

Plus, we know that young people have become reluctant to answer their cell if it’s a number they don’t recognize.

Another issue: Everyone lies. Ask a friend how much money they make and whatever it is, they ad 25 percent. Ask a girl what she weighs, and no matter what it is, she subtracts 12 pounds. Plus, many people are reluctant to talk politics. Maybe they have a different opinion than the people at their office? Or their spouse? Or their community? Or church? Or they have a conscious… but live with a member of the Tea Party.

People also will just tell the pollster what they think the Pollster wants to hear.

And it’s not just the phone. Polls are also conducted on-line which makes it even harder to determine who you’re actually speaking with. I’ve heard about democrats who pretend to be republicans. It’s got a name similar to “Catfishing” – they call it “Pussy Fishing.”

If it’s during an election year, pollsters try and reach “Likely voters.” The problem is that there is no branch of science that can accurately predict future behavior. Unless the behavior you’re predicting is: “Will say something vaguely racist.” And the person doing it is a member of the Trump Cabinet.

That’s because there are two “YOUs.” There’s the current YOU… the one who’s interested in immediate gratification – the one who tells a pollster “Absolutely. Women should be paid equal money for equal work…” And, “Hell yes. Of course I’m gonna vote on election day.”

The problem is: There’s also “FUTURE YOU.”

Future you sleeps late. Drinks too much. Thinks women are over paid… especially when they’re giving lap dances. Future you forgets it’s election day. And it’s not just me saying poll results aren’t reliable, difficult to interpret and easily manipulated… The polling companies that conduct the poll all but says so, too:

Here’s Zogby’s fine print: “Based on a confidence interval of 95%, the margin of error for is +/- 5.7 percentage points.” And Gallup’s is nearly identical.

A recent Gallup poll asked business executives if they approved of President Trump. 34 percent strongly disapproved of the president. Also, it could be as much as 40 percent... or as little as 28 percent... Gallup’s not sure. And those are two very different stories!

Oh, and further undermining our confidence in the outcome, the polling company also makes up a math formula that will add or subtract people’s votes based on national demographics, gender, age, race, population density, and the time they called.

So: If poll numbers aren’t reliable or even the opinions of a representative sample of voters… then WHY THEY FUCK DOES THE NEWS KEEP TELLING US THE RESULTS?

Because there isn’t enough news. And they need to tell voters something. It’s not like you can turn on CNN and Anderson Coopers gonna say: Remember yesterday? Yeah. Things are the same. Just like the day before that, too. That shit doesn’t sell soap. Or diapers.

Jon Hotchkiss is a 14 time Emmy nominee for Politically Incorrect and Bullshit with Penn & Teller. He also hosts Be Less Stupid - which you can see HERE.

This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.
CONVERSATIONS